**2.1. Mass die-offs in marine mammals and discovery of new morbilliviruses**

Since the late 1980s, many mass die-offs have been reported around the coasts of Europe and the USA (Table 1). Approximately 18,000 harbor seals and several hundred grey seals (*Halichoerus grypus*) were found dead on northern European coasts in 1988–1989. The distemper-like gross observations suggested that a morbillivirus could have been the causative agent [6]. Detailed serological, virological, and immunohistochemical examinations showed that the agent was a new member of the genus *Morbillivirus*, named phocine distemper virus (PDV) [5,24]. In 2002, PDV again killed at least 21,000 seals inhabiting the same locales [25,26]. About the same time as the first outbreak on northern European coasts, in 1987–1988 the deaths of approximately 18,000 Baikal seals were reported in Lake Baikal which showed clinical signs identical to those reported in European seals [13]. However, subsequent genomic characterization revealed that the cause of the mass dieoff of Baikal seals was CDV [27-31]. CDV also induced another mass die-off among Caspian seals (*Phoca caspica*), in which many seals died in 1997 and 2000, near Azerbaijan on the western shores of the Caspian Sea [32,33].


CMV, cetacean morbillivirus; CDV, canine distemper virus; PDV, phocine distemper virus.

**Table 1.** Mass die-offs of marine mammals caused by morbilliviruses.

184 New Approaches to the Study of Marine Mammals

Morbilliviruses propagate primarily in lymphoid tissues and induce acute disease. They are usually accompanied by lymphopenia and immunosuppression, which often lead to secondary, opportunistic infections in the host. The distemper viruses, CDV and PDV, often invade the central nervous systems of their hosts, although acute encephalitis is not common in other morbillivirus infections [4]. A notable feature of morbilliviruses is their high host specificity. The natural host of MV is humans, but it can also infect monkeys. Ruminants are the targets of RPV and PPRV. RPV mainly infects cattle, while PPRV infects goats and sheep. Although these viruses have multiple host compatibilities, they induce more severe disease in the primary hosts than in others [11,12]. The natural host for CDV is dogs, but ferrets (*Mustela putorius furo*) have been used as an experimental model due to their high sensitivity to CDV. Recently, the host range of CDV has been shown to be wider than previously thought and expanded to include other wild carnivores, such as Baikal seals (*Phoca sibirica*) or lions (*Panthera leo*) [13-15]. PDV and CMV have been isolated only from seals and cetaceans, respectively. While no morbilliviruses have been isolated from sirenians, serologic evidence of exposure to morbillivirus was reported in manatees

The cellular receptor of a virus is one of the major determinants of host specificity and tissue tropism. The signaling lymphocyte activation molecule (SLAM) has recently been shown to be the principal cellular receptor for morbilliviruses in humans, cows, and dogs [18,19]. SLAM itself was first discovered in 1995 as a novel receptor molecule involved in T-cell activation [20]. It is expressed on various immune cells, such as thymocytes, activated T and B cells, mature dendritic cells, macrophages, and platelets [21,22]. SLAM is also a marker for the most primitive hematopoetic stem cells [23]. The distribution and function of SLAM are consistent with the cell tropism and immunosuppressive nature of morbilliviruses. This indicates that the host range of morbillivirus may be explained by key amino acid residues

In this chapter, we review morbillivirus infection in marine mammals and its possible primary receptor in the host, SLAM. Further, we discuss host–virus specificities based on three-dimensional models of SLAM and risk assessment of morbillivirus infection in marine

**2. Morbillivirus infection and its impact on marine mammals** 

**2.1. Mass die-offs in marine mammals and discovery of new morbilliviruses** 

Since the late 1980s, many mass die-offs have been reported around the coasts of Europe and the USA (Table 1). Approximately 18,000 harbor seals and several hundred grey seals (*Halichoerus grypus*) were found dead on northern European coasts in 1988–1989. The distemper-like gross observations suggested that a morbillivirus could have been the causative agent [6]. Detailed serological, virological, and immunohistochemical examinations showed that the agent was a new member of the genus *Morbillivirus*, named phocine distemper virus (PDV) [5,24]. In 2002, PDV again killed at least 21,000 seals

(*Trichechus manatus*) without showing clinical signs of disease [16,17].

of SLAM on the interface with morbillivirus.

mammals.

The first evidence of morbillivirus infection in cetaceans was described in several stranded harbor porpoises with pathological changes on the Irish coastline in 1988 [7]. A new morbillivirus was isolated and termed PMV for "porpoise" [34]. Since 1990, a severe mass die-off began to affect the striped dolphin (*Stenella coeruleoalba*) population on the Mediterranean coast of Spain and rapidly spread throughout the western Mediterranean Sea, including the coasts of France, Italy, Greece, and Turkey [8,35] (Table 1). A new virus, named DMV for "dolphin" was isolated as the causative agent [9]. A retrospective serologic investigation indicated that PMV and DMV were the agents responsible for another epidemic in bottlenose dolphins (*Tursiops truncatus*) along the Atlantic coast of the USA, for which the causative agent had been initially thought to be brevetoxin produced by a marine dinoflagellate (*Ptychodiscus brevis*) [36-38]. PMV also induced a die-off of bottlenose dolphins in the Gulf of Mexico during 1993–1994 [37-39]. Molecular biological analyses of these cetacean morbilliviruses showed that their gene sequences were similar [9,10]. Based on the similarities, it was proposed that these cetacean morbilliviruses be classified as a single species called CMV.

Thus, morbillivirus infection has a strong impact on populations of marine mammals, as listed in Table 1. In addition to mass die-offs, many smaller-scale die-offs were reported in various oceans. Even if the scale is small, outbreaks of morbillivirus infection can have serious consequences for marine mammal populations, especially among endangered species at risk of extinction, such as Mediterranean monk seals (*Monachus monachus*). In 1997, approximately 50% of the population of Mediterranean monk seals residing along the coast of Mauritania in Africa died suddenly. Morbilliviruses were isolated from the dead seals, although distemper-like lesions were not detected in the animals [40]. Hence, it remains unclear whether morbillivirus was the agent responsible for the dramatic deaths. The involvement of an algal bloom was also suggested as the primary cause [41]. In any case, when populations of many marine mammal species are decreasing, morbillivirus infection may cause a fatal blow.

Host-Virus Specificity of the Morbillivirus Receptor, SLAM, in Marine Mammals:

the well-known 2B4 (CD244), Ly-9 (CD229), NTB-A, and CD84 [49]. All of the SLAM family members have an extracellular region composed of a membrane-distal immunoglobulin variable (V) domain and a membrane-proximal immunoglobulin constant-2 (C2) domain, along with a cytoplasmic region bearing multiple tyrosine-based switch motifs (ITSMs) that bind cytoplasmic Src homology-2 (SH2)-containing proteins such as SLAM-associated protein (SAP) [50]. Evidence is accumulating that the interaction between the cytoplasmic region of SLAM and SAP family molecules mediates a switch to positive or negative signaling in immune cells and plays a crucial role in multiple immune regulations [22]. Genes for the SLAM family receptors are located within a ~400-kb cluster on chromosome 1 in humans and mice [51]. This gene location, coupled with the conserved exon-intron structure of SLAM-related genes, implies that they were generated by the sequential duplication of a single ancestral gene. A SLAM family receptor forms a homophilic dimer by weak binding between the V domains and acts as a self-ligand, suggesting that the receptors can trigger homotypic or heterotypic cell–cell interactions [52]. The V domain of SLAM (CD150) also provides an interface for binding with morbilliviruses [53]. The viral H protein has a strong affinity for the V domain of SLAM, which is 400-fold higher than for self-ligand interaction [54]. The interaction between the viral H protein and SLAM V domain is the initial event in infection with morbilliviruses. The results of recent detailed structural studies have suggested that the interaction changes the microenvironment of the interaction zone for the fusion activity of the F protein, although the mechanism of membrane fusion

mediated by the F protein is not fully understood [55-58].

cultured human kidney cells.

for wild-type morbillviruses.

**3.1. History of the discovery of the morbillivirus receptor** 

The human CD46 molecule was first identified as a cellular receptor for Edmonston vaccine strains of MV [59,60]. The Edmonston strain was isolated from the blood and throat washings of a child with measles using primary human kidney cells in 1954 [61]. It was later adapted to chick embryo fibroblasts and is being used as an attenuated vaccine [62]. This strain grows well in many cell lines, such as Vero cells, and has become the most extensively studied MV strain in the laboratory. However, because CD46, a complement-regulatory molecule, is expressed on all human nucleated cells, its ubiquitous distribution cannot explain the lymphoid tropism of MV. At present, CD46 is thought to be a specific receptor of the Edmonston strain, which is presumed to acquire the ability to use CD46 by adapting to

On the other hand, many wild-type strains have been isolated from clinical samples using the marmoset B cell line (B95a) [63], but they do not grow on many CD46+ cell lines. In order to identify the receptor for wild-type MV, functional expression cloning of a cDNA library of B95a cells was carried out using the VSV pseudotype system. SLAM was shown to be a cellular receptor for wild-type MV [18]. CDV and RPV were also shown to use canine and bovine SLAMs for entry into host cells [19]. Thus, SLAM is thought to be the major receptor

Recently, Nectin 4, a cellular adhesion junction molecule, has been identified as the third receptor for MV in polarized epithelial cells [64,65]. Infection experiments in monkeys

Risk Assessment of Infection Based on Three-Dimensional Models 187
